Dimensionality reduction (DR) techniques are fundamental to the analysis of high-dimensional data, yet their substantial computational cost limits their use in interactive settings. To address this limitation, we present GPU-accelerated Progressive Uniform Manifold Approximation and Projection (GPUMAP), an efficient progressive DR algorithm designed for interactive use. We identify three atomic operations within UMAP and restructure them into a progressive pipeline. As part of this adaptation, we introduce modifications to UMAP, including operation queues, instance-wise local join, and scheduling strategies. We further accelerate computation by leveraging GPU parallelism, enabling interactive analysis of million-scale datasets on commodity GPU hardware. Finally, we employ our implementation to systematically study the behavior of progressive DR, investigating how projections evolve over time relative to the ground-truth projection and how biased data streams influence the quality and stability of the resulting projections.
Progressive Visual Analytics (PVA) systems have demonstrated the effectiveness of time-bounded, incremental computation for exploring large-scale data. However, our survey reveals that existing PVA implementations are often difficult to reuse and extend, as their progressive behaviors are typically realized through ad hoc, hard-coded mechanisms distributed across servers and clients. To address this limitation, we present ProGram, a grammar for progressive computation pipelines. Informed by a comprehensive analysis of prior concepts and design practices in existing PVA systems, ProGram is designed to support core progressive features while reducing implementation burden. Furthermore, we extend the design space by introducing novel mechanisms, such as lockstepping, forking, and temporary prioritization, to better support diverse usage scenarios. Finally, we demonstrate the expressiveness and coverage of ProGram by reproducing and extending existing progressive visualization systems, presenting a gallery of progressive computation pipelines, and discussing the grammar coverage.